Research map / Probabilistic and scientific modeling / Molecules and proteins
Interpretable Discovery of One-parameter Subgroups: A Modular Framework for Elliptical, Hyperbolic, and Parabolic Symmetries
Pavan Karjol, Vivek V Kashyap, Rohan Kashyap, Prathosh AP
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace Augerino augmentation-based discovery with structural invariant representation (elliptic) | HGNet vs Augerino · Double Pendulum | prediction error (MSE): (6.2 ± 2.9)×10^-4 -> (2.0 ± 0.1)×10^-4 | Improved |
| replace Augerino with HGNet (elliptic) | HGNet vs Augerino · Invariant Polynomial u(x) | invariance error: (7.8 ± 4.3)×10^-2 -> (0.4 ± 0.2)×10^-5 | Improved |
| replace Augerino with HGNet (elliptic) | HGNet vs Augerino · Invariant Polynomial u(x) | prediction error (MSE): (1.7 ± 1.4)×10^0 -> (1.0 ± 0.0)×10^-5 | Improved |
| replace Augerino with HGNet (elliptic) | HGNet vs Augerino · Angled Sine | cosine similarity: 0.616 ± 0.075 -> 1.000 ± 0.000 | Improved |
| replace Augerino with HGNet (elliptic) | HGNet vs Augerino · Angled Sine | prediction error (MSE): (3.4 ± 4.7)×10^-2 -> (1.1 ± 1.2)×10^-2 | Improved |
| replace Augerino with HGNet (elliptic) | HGNet vs Augerino · Double Pendulum | cosine similarity: 0.908 ± 0.102 -> 1.000 ± 0.000 | Improved |
| replace Augerino with HGNet (elliptic) | HGNet vs Augerino · Top Quark Tagging | accuracy (%): 74.9 ± 2.4% -> 83.0 ± 0.1% | Improved |
| use hyperbolic (boost) invariant representation instead of elliptic (ablation) | HGNet · Top Quark Tagging | accuracy (%): 83.0 ± 0.1% -> 84.4 ± 0.4% | Improved |
| use hyperbolic instead of elliptic invariant representation (ablation) | HGNet · Top Quark Tagging | cosine similarity: 1.000 ± 0.000 -> 0.996 ± 0.003 | No clear effect |
| replace Augerino with HGNet (elliptic) | HGNet vs Augerino · Top Quark Tagging | cosine similarity: 0.994 ± 0.003 -> 1.000 ± 0.000 | No clear effect |
Closest papers by meaning
- Explicit Discovery of Nonlinear Symmetries from Dynamic Data (ICML 2025 · similarity 0.88)
- AtlasD: Automatic Local Symmetry Discovery (ICML 2025 · similarity 0.87)
- Learning Infinitesimal Generators of Continuous Symmetries from Data (NeurIPS 2024 · similarity 0.87)
- Discovering Symmetry Breaking in Physical Systems with Relaxed Group Convolution (ICML 2024 · similarity 0.86)
- Lie Algebra Canonicalization: Equivariant Neural Operators under arbitrary Lie Groups (ICLR 2025 · similarity 0.86)
- Lie Group Decompositions for Equivariant Neural Networks (ICLR 2024 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
